Smart Healthcare, Biomedical Signal Processing, VLSI/Embedded Systems, Healthcare Cyber-Physical Systems, Security of CPS, VLSI DSP Systems,
134
Scopus Publications
Scopus Publications
Denoising of wrist PPG using SADWT technique for wearable devices Sakshi Tiwari, Surendra Kumar Saini, Amit Mahesh Joshi, Ravi Kumar Maddila Proceedings of SPIE the International Society for Optical Engineering, 2026 Photoplethysmogram(PPG) signal observed various noises and motion artifacts during recording from the biosensors in the wearables devices. For health monitoring, physiological parameters detection is valuable for indicating the cardiovascular status of a patient. This work exploits novel approach for acquiring denoised PPG signals acquired from wrist. By employing multiresolution analysis, the PPG signal is segregated into various subbands through the Discrete-Wavelet Transform (DWT), in which sum of absolute differences(SAD) filtering scheme is applied on the low frequency subband. After removal of motion artifacts, popular distortion measures parameter is evaluated to indicate the quality of the obtained denoised signal recorded from wrist of the 34 subjects. The experimental results indicates that the proposed SADWT filtering method outperforms existing techniques, achieving an SNR of 25.78 dB respectively.
Hardware Implementation of Weight Computation Techniques for Pair-Wise Orthogonal Transform Used in Onboard Satellite Spectral Preprocessing Priyank Prajapati, Bandaru Poorna Chandra, Anand Darji, Jimit Gadhia, Ankur Garg, Amit Joshi Proceedings of the National Conference on Communications Ncc, 2026 In space data communication, the design of an onboard hyperspectral or multispectral image compression system is challenging due to limited power, storage, and communication bandwidth. As per the Consultative Committee for Space Data Systems (CCSDS) (120×3g1), the Pair-wise Orthogonal Transform (POT) provides efficient data rates for spectral transformationbased image compression. However, its hardware architecture is yet not explored despite its better compression efficiency due to the design complexity of its weights computation function. The proposed work presents the LUT-based and Computation-based hardware implementation techniques of a weight value derivation architecture, synthesized to higher precision parameters, considering resource utilization constraints. The Computation-based weight derivation uses CORDIC IP and custom division IP. The hardware implementation of the Computation-based weight value derivation architecture on the KCU105 FPGA board achieves only 11.61% LUT utilization for the highest-precision parameter. Further, the total delay and maximum power consumption are observed to be 3.785 <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$n s$</tex> and 1.174 <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$W$</tex> at an operating frequency of 150MHz, respectively. The architecture of POT is pipelined to achieve 1 clock per 2 pixels of processing. The proposed hardware implementation outperforms existing work in terms of precision, area efficiency, and throughput.
Semi Supervised Cyber Attack Detection System for Smart Grid Richa Sharma, Amit M Joshi, Chitrakant Sahu, Gulshan Sharma, K. T. Akindeji, Sachin Sharma Proceedings 30th Southern African Universities Power Engineering Conference Saupec 2022, 2022
A survey on recent advances in speech compressive sensing U. P. Shukla, N. B. Patel, A. M. Joshi Proceedings 2013 IEEE International Multi Conference on Automation Computing Control Communication and Compressed Sensing Imac4s 2013, 2013